The United Nations now says its 2009 headline-grabbing announcement that 1 billion people in the world were hungry was off-target and that the number is actually more like 870 million.

The U.N. Food and Agriculture Organization blamed flawed methodology and poor data for the bum projection, and said it now uses a much more accurate set of parameters and statistics to calculate its annual estimate of the world's hungry.

FAO issued its 2012 state of food insecurity report on Tuesday, and its core point was to set the record straight about the number of the world's undernourished people, applying the more accurate data retroactively back to 1990.

And the good news, FAO said, is that the number of hungry people has actually been declining steadily -- rather than increasing -- over the past two decades, although progress has slowed since the 2007-2008 food crises and the global economic downturn.

FAO said that if the right action is taken now to boost economic growth and invest in agriculture, particularly in poor countries, the U.N. goal of reducing by one-half the number of the world's hungry people by 2015 is very much within reach.

To be sure, 870 million hungry people is still far too many hungry people, said the heads of the three U.N. food agencies in a forward to the report.

"In today's world of unprecedented technical and economic opportunities, we find it entirely unacceptable that more than 100 million children under the age of five are underweight, and are therefore unable to realize their full socio-economic and human potential," they wrote.

FAO made headlines in 2009 when it announced that 1 billion people -- one-sixth of the world's population -- were undernourished. A high-level summit was called at FAO headquarters in Rome. The U.N. chief went on a daylong hunger strike to show solidarity with the 1 billion. The Group of Eight devoted much of its summit that year to pledging $20 billion for seeds, fertilizers and tools to help poor nations feed themselves.

It turns out, though, that the projections were wrong. They were calculated using figures from non-U.N. sources that were fed into the U.N.'s number-crunching model, because FAO was under pressure from governments to quickly come up with an estimate of how many people might go hungry from the dual crises of high food prices and the global downturn, said Kostas Stamoulis, director of FAO's agricultural development economics division.

"There was considerable fear that that combination of lower incomes and higher prices was going to cause significant undernourishment," said Jomo Kwame Sundaram, FAO's assistant director-general for economic and social development.

But now, "no one really knows for sure if at any particular period whether that 1 billion figure was actually reached or not," he said, explaining that the goal is to assess chronic hunger, rather than spikes caused by temporary food shortages and price hikes.

What the U.N. couldn't know at the time was how well governments would respond to the crises to protect their poorest, and how individual families would make sacrifices -- such as in health care or education -- to make sure they had enough food on the table, said Jomo.

"So it was in a way a mistake, an error, that we have made with many other organizations at the time," said Pietro Gennari, the statistics director for the FAO's economic and social development department.

Already, in 2010 FAO said the number of the world's hungry was down to 925 million, though it didn't explain how it came to that figure. On Tuesday, FAO officials said the new number of hungry people for 2012 was 870 million, and the organization revised all of its figures from 1990 using a new methodology based on:

--New population data from the U.N. Population Fund. China's population estimate for the 1990s, for example, has been revised upward by some 25 million people, whereas Bangladesh's population has been revised downwards by 17 million people. That affects hunger statistics because food production figures are divided by population figures.

--Estimates of food loss at the retail level. Previously, FAO considered the amount of food produced as the amount of food available to feed the world. The agency didn't take into account that one-third of all food produced is wasted along the distribution chain, either because it spoils, is eaten by rodents or is otherwise inedible. The 2012 survey takes into account these losses.

--New demographic and health surveys that measure people's height. FAO had been relying on 20-year-old World Health Organization statistics to determine how many calories were needed based on body mass, with taller people requiring more calories than shorter ones. For the 2012 report, FAO is using new height surveys to determine caloric requirements.

"What we are saying is we are recalculating everything with new data, improved data, and what we believe to be an improved methodology," said Jomo, the assistant director-general.

That said, he stressed that all hunger estimates by their nature are conservative. FAO's caloric requirements, for example, assume a sedentary lifestyle, even though many of the world's hungry often do strenuous manual labor, thus requiring more calories to meet their food needs.

And the figures only look at calories needed for energy, not the protein and other nutrients that are critical for development, particularly for the young.

Oddly enough, with the new number-crunching methods, FAO discovered that the world's hungry actually did hit the 1 billion mark, but it was back in 1990-1992. The world just didn't know it then because the FAO was using the old data that set the hungry figure for that period at 848.4 million.